Alexander Talei is the Founder of Obermind, a fast growing quantitative research and technology development firm which has set out to help financial technology and trading firms to build, deploy and scale on its new platform. We asked him to tell us more about this venture and the toolsets and robust infrastrature Obermind is building for intelligent automation across a wide array of markets.
Why did you decide to engineer the Obermind platform and what factors influenced how you went about that?
Simply because there was no platform out there that had all I needed and did all that I wanted to do. To get what I wanted I would have to put together some very expensive tools that I frankly couldn’t afford. So I called a friend of mine, just one phone call and we put together our team in a week’s time and started building the Obermind Platform. Most other firms, not just startups, feel the same way. The tools out there are insanely expensive and so closed-ended that no one really controls their own technology. People are locked with off-the-shelf products; they can’t stand-out or get new ideas out in the market without building their own systems from scratch, which is not the cheapest endeavour out there. I always knew that “the platform” didn’t exist – so we built it.
In what ways have you leveraged the latest technology to ensure the platform has maximum flexibility and configurability?
Well for us it was easy to build the platform for speed and performance. That’s what we have been doing all our adult lives, but the challenge was: how can we architect the platform so we maintain performance and agility without cutting back on performance or locking ourselves into a box and ending up being another legacy system? How can we make sure our platform is future-proof? Should it support both decentralised and centralised use cases? How can we build a system that is built to perform for a prop trading firm and for a broker/dealer distributing consumer-facing applications across numerous devices to thousands of end-users?
It was a big challenge and I knew we needed to make sacrifices somewhere to reach this goal, kind of forking out to suit a specific user base or user case only. But I felt we would become like every other platform out there then, that we would have a product instead of an actual platform that our user-base could take on and build out to things we ourselves couldn’t even imagine. Most who have launched financial technology understand the wariness you get when it comes to introducing new code, features or updates. That’s what prohibits progression in one way for most firms, I didn’t want to end up in that box.
It was a difficult decision to make, but I decided that we will not give up on the platform’s interoperability and performance. So we ensured that the platform:
- Is portable and cross-platform.
- The APIs and libraries are split into core and non-core, while remaining open and accessible to all our users.
- The API must support numerous use cases and support popular scripting languages (like Python for back-testing) as well as OOP languages to build more high performance systems and not be “just another REST API” – as that kind of shuts down most production use cases.
- Instead of creating a “product”, we need to build applications on the platform and distribute it for free to our clients who don’t have in-house developers and want to launch and run their platform with no-code.
- Enabling users to actually own their platform and develop or configure the systems they want to launch to a greater extent than what is possible today. We give our clients the ability to branch out on their own.
- Be transparent about it all, constantly thinking “I need to keep this proprietary, I need to ensure this can get monetised” will cripple usage, adaptation and progression. Just put it out there, if it’s good – I’m sure others will copy, that’s just how things work. But at least we will continue to progress and be the platform we want to be.
This way, I believe we can maintain a future-proof platform that is open to a wide array of use cases without jeopardising the platform’s ability to scale into the future and our ability to continuously roll out new features and applications. Most importantly, every user has the ability to branch out at any time and build what they want using core libraries while cherry-picking new functionality using the API as we move ahead.
What key features and functionality are available with Obermind?
Let me run through a few use-cases that your readers might be interested in briefly, that should answer the question easier:
Sell-side bridge / connectivity hub: We have built libraries and have a fast growing API connectivity list enabling firms to manage their entire connectivity network from one platform. From market data to order execution, routing and beyond. The logic and rules can easily be hard-coded by users or they can use our pre-built order execution algorithms and easily update parameters for no-code users.
Multi-asset buy-side portfolio manager: For a quantitative portfolio manager, we enable coding, back-testing, optimisation, simulations, quant libraries and more to devise their strategies. Once ready to deploy, strategies can easily be distributed on numerous venues across a global cloud as high-performance cross-platform console applications. Most importantly, all of this can be monitored and controlled through feature-rich, pre-shipped or customised user interfaces where literally every messaging, regardless of asset class and venue, is unified. For a high-touch portfolio manager, well it’s even easier. We have built it pretty much to be “plug-and-trade”.
Broker-dealers and Exchanges: For broker/dealers and exchanges, we have an end-to-end solution, where they can distribute front-ends for native Windows, web and mobile, completely custom made and owned by them so they can create new features, design and configure their own applications without us – or simply just “plug-and-trade”.The Obermind Platform has modules to manage anything from on-boarding and security (OAuth, multi-factor authentication) to multi-asset credit management (using our wide variety of pre- and post-trade credit engine tools capturing trades in microseconds), to managing payments using our online web application and more.
Firms can then easily serve market data and manage order execution using our low-latency Stream server, easily deployed on any Windows or Linux server in minutes. It comes as both native portable applications running over TCP for high throughput using our binary protocol, or via built-in FIX server that can be configured and deployed in a day and as a websocket application for broker/dealers redistributing over the internet.
This can easily be configured with no-code to run both asynchronous and synchronous risk-management and internalisation or build light-weight sell-side algorithms using our sell-side library with a few lines of code to automate building limit order books, order execution, pricing or internalisation.
Centralised, decentralised – completely up to them. Our proprietary in-memory financial data server ObermindDB is being used in both “worlds” and can be configured to suit numerous use cases. They can then plug that into standard pre-configured SQL RMDBs or use custom user-specific time-series databases or proprietary systems.
Our suite of data visualisation tools can be highly configured to display anything from decay curves, hit rates, fill ratios and more.
Fintech: Other than creating slick cross-platform applications with our pre-built apps using Razor and React, or using their own framework and integrate Obermind libraries, they need connectivity, they need to be able to manage settlements, payments and more – that’s where Obermind really helps out launching solutions much faster and easier than before. Instead of rewriting and reinventing the wheel, the platform has all the modules they need built-in so they just focus on building their products and solutions and getting to market faster than their competitors.
Proprietary Trading Firms and Prime Brokers: Obermind is used by prop firms to manage centralised credit, risk and connectivity gateways which traders connect via to prevent run-away algos or to customise limits, connectivity, market access and much more – and all built for a multi-asset world. We built it to suit both low-touch algorithmic and high-touch firms.
How would you describe the key benefits of Obermind and why it is different from many other trading platforms?
It really is a platform to use to build new platforms and financial systems on top of. Our pre-built applications can also be “white labelled” and configured to launch a platform in days for those who don’t have the resources to develop their own system.
Other than the technical aspects I briefly described, more importantly we want our clients to own their platform and build value instead of just paying license and ticket fees. So we do provide source-code, redistribution rights and one-off perpetual licenses, so firms know what their system costs without any surprises and they understand what value they are building.
We built the platform, we probably know it better than most, so we are always there on Slack, email or on the phone to get things done, whether it is custom development or helping configuring the platform or trouble-shooting, we’re there every day of the week.
Every client gets a dedicated Project Manager that follows through the entire project life-cycle from planning and requirements to production and updates. We use agile project management tools, with full audit-trail of requirements, work and communication to ensure the firms we work with and their stakeholders, management, compliance and developers alike get the tools they need to understand the entire project life-cycle.
What we have seen for now that attracts most, is that we don’t have any ticket fees, per million fees, or if launched on our managed cloud instead of on-premises, there are no licensing or software re-distribution fees either. This dramatically reduces technology and transaction costs from the traditional ~6-8% of revenue that most financial firms are exposed to. I do believe in the long-run, this will trickle down to lower end-users transaction costs as well.
What types of trading firm and FX market participant are you targeting with the new platform?
Right now it is mainly prop trading firms, fintechs and broker/dealers that are rolling out using Obermind. Mainly in Futures and crypto trading space but also in FX. I believe more and more FX firms will join us. Most firms have already realised the importance of standing out with their own technology, not just to encourage brand-awareness and create captive audiences but also to take control of their future and be able to launch new functions, features and asset classes as they please and not when everybody else does it. In FX, there is very little that differentiates between providers and competitors, particularly when it comes to the technology offering. I hope Obermind will help change that.
What deployment options are available with Obermind?
Community License, Managed Cloud or On-premises.
Our Community License is for individual and academic use. It has everything needed for the startup or buy-side “bedroom-traders” to research and devise automated strategies and start out their portfolio management business or fund. That’s free. We also provide free hosting for it as long as we can afford it.
For trading firms and institutions we have two options:
Managed Cloud, where the platform is hosted on our hybrid network. There are no additional fees other than the regular hosting fees. No licensing, software, plugin or ticket fees or such. We cover more than 17 locations from near proximity to co-location and take care of cross-connects, security, optimisation and everything else in-between. For firms who want a more proactive management or take care of the management on their own, we do have the option to just monitor up-time and they do the rest.
We set up custom SLA’s when needed and support a real hybrid cloud infrastructure for financial firms, from latency sensitive usage to low-cost storage or container needs on the “big three” clouds.
Our On-Premises licensing is simple and customisable, we can either do a one-off perpetual license, source-code license or simple monthly subscriptions – whatever our clients prefer. We also don’t charge ticket fees with On-Premises. Our clients can then host the platform wherever they want, they still get our full support all the way.
Whatever option our clients choose, we believe that training is crucial to get out what you need from the Obermind Platform. We provide on- and off-site training and certification when required to ensure the platform is used to its maximum potential.
What plans do you have to roll out more functionality with the platform over the coming months?
Right now we are rolling out our free market data API. Millions of historical data points, near real-time market data on almost 30,000 instruments and markets, from ETFs and equities to FX and crypto – to statistical indicators, dividend calendars, SEC filings, commitment of traders reports and much more. Free to use, super-cheap to redistribute. Our API and integrations list is constantly growing, so worth checking out there. There is more in the pipe-line but too early to mention.
What’s the easiest way for people to find out more about Obermind?
Our website https://obermind.com has most of it covered. We are updating it quite frequently and adding documentation and more. We are also pushing out more of our pre-built applications in stages for free usage.